Abstract: Biological systems often display sudden, radical shifts in their state in response to small changes in the environment. Such tipping points include disease outbreaks, the collapse of populations, and cellular differentiation. Given that many of these tipping points have huge health implications-either positive or negative-it is essential that we develop an increased understanding of these transitions. Although a complete understanding requires detailed knowledge of the system, recent progress in the theory of nonlinear dynamics argues that there are generic early warning indicators preceding such tipping points. In particular, theory predicts that fluctuations of the system should become larger and slower as the tipping point is approached, potentially signaling dramatic changes in the biological system. Here we propose to measure these early warning indicators experimentally, first at the level of the population before sudden collapse and next within individual cells before a change in cell state. We will start by using laboratory microbial populations that display catastrophic collapse in deteriorating environments. The primary model system will be budding yeast, which cooperatively breakdown the sugar sucrose. The cooperative nature of this growth means that in each environment there is a critical population size required for population survival; in response to a deteriorating environment the population therefore collapses suddenly. However, by measuring the day-to-day fluctuations in the population size we have observed both an increase in the size and timescale of the fluctuations as the tipping point approaches. To our knowledge, this preliminary work is the first experimental measurement of these early warning indicators in a population that displays catastrophic collapse. Now that we have demonstrated the feasibility of the approach, we will measure these early warning indicators in different deteriorating environments, in the presence of cheater cells (that don't contribute to breaking down the sugar), and in laboratory ecosystems containing more than one species. We will also study population fluctuations before collapse in spatially extended populations; this coupling between neighboring regions is expected to obscure some early warning indicators, but may also cause spatial patterns to emerge before the tipping point. Finally, we will move to the level of the cell and explore whether it is possible to experimentally observe these early warning indicators before a phenotypic switch, initially in sugar utilization. Theory predicts that gene expression will become more variable before important cellular transitions such as differentiation. In addition, the timescale of fluctuations of the key proteins involved in the genetic circuit should increase before the switch. The experiments described in this proposal represent a unique opportunity to explore the universal signatures of biological systems before sudden transitions. Public Health Relevance: Biological systems can experience sudden tipping points such as a disease outbreak, population collapse, cellular differentiation, or cancer metastasis. As many of these transitions have profound implications for human health, there would be extreme value in being able to determine in advance when such transitions are about to take place. Here we propose to study possible early warning indicators of impending tipping points that are based upon the fluctuations of the biological system.